INDEX

Explanations

structured data or lists within a text

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 14-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L14R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.14.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

phalt

-0.16

Ð·ÑĮ

-0.15

SOLE

-0.15

_here

-0.14

ecek

-0.14

 Handler

-0.14

Ã¢ng

-0.14

Ð°Ð»Ð¸

-0.14

Ð»ÐµÐ²

-0.14

sted

-0.14

POSITIVE LOGITS

gil

0.15

unsch

0.15

namen

0.15

chl

0.14

 Lane

0.14

Ãºsqueda

0.14

ãĥĥãĥĪ

0.14

 slim

0.14

_helpers

0.14

.band

0.14

Activations Density 0.000%

structured data or lists within a text

No Comments

No Known Activations